- From: Jonathan Borden <jonathan@openhealth.org>
- Date: Wed, 12 Feb 2003 18:45:09 -0500
- To: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>
- Cc: <www-webont-wg@w3.org>
Peter F. Patel-Schneider wrote: > > Because they can be applied to classes and properties. If they can't, then > they are no different from data-valued properties. If they can, then they > are are completely different category. Sure, let rdf:comment be applied to classes and properties, it is a (data valued) property *like any other* > > That said, I'm not in favour of syntactically tagging properties that show > up in annotations as annotation properties. This view has the benefit of > not producing any change in OWL Full. > What is the fundamental harm in allowing :foo rdf:comment "anything" . where :foo might be any of a class, property or ontology. Now if this causes issues for ontology versioning etc, so be it -- that is a function of the system we've designed. The bottom line is that one could say that if someone changes a comment, the whole ontology is changed. Or one might choose to compare ontologies, or any other individuals, without talking into account this, or any other property. I fail to see why this ought to be a fundamentally special case. I don't see why this annotation issue is such a fundamental problem that it deserves special semantics. Strings attached to graphs are just that. Jonathan
Received on Wednesday, 12 February 2003 19:07:57 UTC